Q: You run a shell script as the ENTRYPOINT that spawns multiple background worker processes. When you `docker stop` the container, it always takes exactly 10 seconds (the timeout) before stopping, and the workers don't clean up properly. What is the root cause?

When Docker sends SIGTERM via docker stop, it delivers the signal to PID 1 inside the container. If PID 1 is a shell script (/bin/sh or /bin/bash), the shell does NOT forward signals to its child processes by default. The shell itself ignores SIGTERM, so nothing happens for 10 seconds until Docker sends SIGKILL.

• Use exec to replace the shell with the main process: the last line of your entrypoint script should be exec ./my-worker instead of ./my-worker. This makes the worker PID 1 and it receives SIGTERM directly.
• Trap signals in the shell script if you must manage multiple processes:
• Use --init flag (docker run --init) to inject tini as PID 1, which properly forwards signals and reaps zombie processes.

*Fixes:*

#!/bin/bash
trap 'kill $(jobs -p); wait' SIGTERM SIGINT
./worker1 &
./worker2 &
wait